On the final day of the Roundup for 2012, Gunnar Hillert, who organizes DevNexus in Atlanta, suggested that he and I convene a session on Organizing your own Conference. Since I had previously done such a session in March 2008 with Stephan Janssen (of DevOxx): Java Posse Podcast #197, I was initially reluctant. But open spaces is about seeing what will happen, and there were others to learn from. In addition, I’ve learned a lot since then with Codemash. We decided to go for it.
The two conferences couldn’t be more different, in spite of the fact that they are both developer-organized. CodeMash attracts a diverse group of developers, intent on learning other technologies in order to improve what they do with their own. DevNexus focuses mainly on Java developers. CodeMash is in the sleepy town of Sandusky, OH, meaning that pretty much everyone has to drive (or fly) to get there. DevNexus is in the heart of downtown Atlanta, which draws an audience from the metropolitan area.
We talked a lot about what it takes to organize a conference. I think that the real key is that there is someone who is willing to put his or her neck out without knowing whether or not it will be a success. Signing a catering or rental agreement with a venue is a scary thing and I’ll be forever grateful to Jim Holmes for taking on this risk for CodeMash.
Other user-organized conferences that people might be interested in checking out:
That Conference
Strange Loop
PyCon
Stir Trek
After the Organizing your own Conference session, I joined Andrew Harmel Law in a discussion on Variations in Pair Programming. He had originally intended it to be a lightning talk, but 5 minutes is really short! When he told me a little about the rules his company had come up with around Pair Programming, I immediately challenged one of his tenets: that pair programming should be 2 people, 1 keyboard, 1 mouse. We’ve been effectively programming with 2 people, 2 keyboards, 2 mice, 2 monitors, facing each other rather than sitting next to one another, for a few years now. He was intrigued by the variation, and we convened a session to see what others thought.
Like both of our companies, many of the participants said that they do not pair program 100% of the time. The developers decide when and where to pair program. Our programmers have wide latitude around self-organization. Some teams have a dedicated “pairing computer” in addition to their own laptops. Andrew was thinking that he might take that back to his company as well. I’ll be intrigued to see if he does. And I can’t wait to expose our team to Andrew’s Tarot Cards of Pair Programming.
Others expressed concerns about teams of various skill levels in pair programming. Mainly these concerns seem to have come from people who were thinking about pairing, rather than those who were actively doing it. Those who actively pair program see it as a way to bring a new team member up to speed on a project quickly, and also as a way to take advantage of the individual strengths of team members to build a better overall solution. It was a fun session, and it should be interesting listening once it becomes available on the podcast.
The last session of the conference was a wrap-up session where we all talked about what we got out of the conference and what improvements can be made. This is the 6th Java Posse Roundup and it became obvious to many of the old-timers that the newbies are becoming integrated into the group more quickly. Apparently, they just have more mentors. It was also interesting to learn that at least 2 of the attendees this year had never heard of the podcast but decided to come to the conference anyhow.
On Friday night, many people head home. Those of us who have attended in previous years tend to stay until Saturday, because flights on Friday often mean missing the last session. This has led to a variety of experiments around Friday night activities. For the past several years, a group has gone up to Mount Crested Butte to a very nice (star-rated) restaurant called (appropriately for this crowd) Djangos (Geek humor alert: There’s a python web framework called Django). They have small plates and wonderful wine. The groups have tended to order “One of Everything” which is a great way to try their wonderful food. But we also found out that we could rent a yurt, accessible only via nordic skis or snowshoes. We didn’t have time to organize it last year, but this year, a group expressed interest and we were able to organize catering at the yurt for a group of about 25 people (20 others went to Djangos). This was a fabulous experience, in spite of the bad trail conditions. We hiked out to the yurt, and had a wonderful meal catered out there, complete with tiramisu for dessert. It was so much fun that it seems likely to continue in subsequent years.
People filtered out of town throughout the day on Saturday. About 30-40 people showed up for a departing breakfast. Barry Hawkins organized it and cooked up endless amounts of eggs while many others chipped in to help, such as cooking bacon on the grill, and making biscuits and gravy. It was a great way to use up the food that people had bought for their week in town, and way more fun and interactive than the smaller group that a restaurant could accommodate.
So now the Java Posse Roundup 2012 has come to a close, but the git repository remains and the projects that were started will continue to be developed. Even more significantly, the knowledge and energy that everyone took away from the conference will spread throughout our companies over the next several months until it’s time for the Java Posse Roundup 2013. I’m planning to be there, of course.